Ever since the Core Web Vitals tool was launched by Google and the major SEO publications began promoting the results as the next ranking factor, Clint was skeptical and publicly said that he didn’t ‘believe’ that it would ever be a ranking factor that was any more significant than page speed itself.
He believes that if your site was very slow and you optimized it for speed and made significant changes that you very well could obtain higher rankings. However, if your website was already “fast enough” and you optimized it, even more, you’d be less like to see any ranking improvement whatsoever.
In this test, he did more on data collection, rewrote the Javascript, came up with some keywords set for algorithm changes and was built to get a wide range of markets,
local, ecomm, national, etc. He correlate each of the different categories to validate if Core Web Vitals is really a Ranking Factor or not.
Test Downloads
Transcription below:
Clint: All right. So it's 43 pages long. I have some here and there just to bust it all out. If you spent any money optimizing for specifically for Core Web Vitals. If you collected money optimizing specifically for Core Web Vitals, you may be kicking yourself in the rear end at the end of this.
If you listen to my advice, months ago when I told you not to be optimized for Core Web Vitals, you're going to be pretty glad that you listened to me because, finally proved it. Core Web Vitals is not used by Google as a ranking factor. Despite what Google said in their announcements, despite what every SEO journal, SEO newspaper, SEO blog, SEO talking head said, Google is not using Core Web Vitals for ranking determinations.
There's only one part that falls into the Core Web Vitals scope. And it's not even something that's announced on the testing tools that correlates with rankings. And you'll kind of figure out why Google pinch Core Web Vitals so bad or so strongly, in order to get SEOs to jump on this gravy train. And then, when they pushed it all out, and we're done with it, at the end of August, nobody saw any significant improvement, unless they did one thing.
Let's go through this, and show you how I got onto this track. This was an adventure. I started in the end of September. As you can see, those dates are not a lie. That's when I started this test. And then I finished it last night, we're writing this up, it took 2 and a half hours.
Marie had the rest of the entire dossier printed and she was still waiting for me to finish this one. So I do currently and in the past offer a PageSpeed optimization service for WordPress. And I do that without any sort of "Oh, you shouldn't do that" Because if you increase the webpage significantly, if you increase the optimization of the say pages loading in 10 seconds, and you get it loading in 3, you will see a rankings boost because that is a significant difference.
If you are optimizing a page that loads in 3 seconds and you wanted to get it and you get it loading it to. You're probably not going to see much of a difference. And to this day, I still turn away business for Core Web Vitals optimization and PageSpeed optimization in general. If your site's just not bad.
If it's not, if it's good, if it's within that range, if it's still loading within 3 seconds, if I can get on there and as a user, I can still click around on a mobile phone and it's just fine. I tell you save your money, put it somewhere else. And that advices seem to work well for me.
There are plenty of people who jumped on a gravy train, though and promoted Core Web Vitals as something to sell to optimize for. You can see here, Josh McAdams. He said he charged customers 1000s. There's 18 people here who voted yes. I charged for specific package. 52% say no. I would argue that, you know, based on the size of the group and the type of people that are in there that this number is probably a lot higher.
I have no doubt that people were showing all those talking heads and showing all those articles. And they're charging customers a lot more for Core Web Vitals because A. It's a ranking factor. It's the latest and greatest thing. We need to be optimized for it. So you should spend me you know, anywhere from $250. I had one client do that too, and other clients who literally spent 1000s of dollars, top demise for this metric.
Here in the group, you see 123 people think that is a vital or an important ranking factor and the 205 said NO. But I will say that I've seen people that voted NO who were talking about it in some context or another saying that you need to do it. So no, but its user. The biggest argument was no, it's not a ranking factor, but user experiences. I bounced rate, pingback etc.
And if you've been a member of SIA for any period of time, you've by now had a chance to go through some of the tests. You see that we've tested those things bounce rate, time on site, etc. And none of those appears to be ranking factors. In all fairness, we probably need to change test those again. But based off of the data that I've got here, I don't know that it would be a valuable use of time or for us, because based off of the data that I've seen that I collected during this month, it's not likely that that's different.But in May 2020, Google announced Core Web Vitals and then in November of 2020, they said, oh, it's going to be a ranking factor. And then it was on SEO is repent misting this thing talking about it everyone was you can turn a page or go to a group or anything without someone asking about a Core Web Vitals optimization. And this number is not a lie.
I do know some businesses that have spent well over that to optimize for Core Web Vitals to get these metrics in the green. Which is pretty sad. So June 15, Google announced that they're rolling this thing out, and they said, it's going to take till the end of August.
I don't know about you guys, but I have not seen one update. In all the years that I've been doing this that took a month and a half to roll out. That just sounded shady to me right off the bat. And as I say, they call me Blackhat. Is that I mean, what is Google thinking? Do they honestly believe that we believe and a lot of people to this day still do believe? Oh, it's not rolled out completely yet. This isn't done.
The only thing that I could say that maybe would cause that timeframe would be a shakeup in the SERPs. And from July or from July all the way till now. There has been no significant shake up the shirts that you could track and Core Web Vitals.
So, Lee Witcher did an episode on SEO Fight Club on the 22nd. They did I got wrangled into doing this. And he put a whole bunch of data, got some keywords and analyzed it. And based off the data, he could find no correlation. Lee's smart dude, and his presentation made it seem like there was some funny math going on there. So I wanted to look at it from from a third party perspective. I want to dig dive deep into it. Kind of duplicate what he's doing.
If I can duplicate it, then I can go to all the naysayers that poopoo Lee's research and say, Look, I did it again. And he's right.
On top of that, not too long after Ted Kubaitis actually found this script. And it's on this page here. When you get the dossier, you can click on that link, it'll take you to a website that Google set up. It's like 5 pages teaching you to install the script. And this script sends us data CLS FID LCP to you Google Analytics account, and then it collects that data for you inside your Google Analytics account, which is pretty simple, where it goes, and you can see that firing off and working inside of the Chrome Inspect tool.
What you'll see are these numbers right here. Once you look very very closely at these numbers, as you're doing this, just kind of looking. See if you see something that Ted and I saw. Wait for it. I'll give you a little bit of a hint.
Random match. Look at these LCP variances, FID, all these variances. CLS kind of consistent, but it's still a big jump there, maybe you can kind of have to read out the caching maybe.But look at these variances, LCP, these are all the same webpage, the same webpage, the same browser, the same computer, the same internet service provider, the same IP.
If you're hitting the same page over and over again, like I told you, all you have to do is you open a page, you click anywhere these to fire off, and then you push, reset, refresh, and then the CLS cups. And those numbers will go over and over again. As long as you keep doing that. It's the same person, the same web browser and etc. You would expect all those numbers to be closer. I mean, you get some latency and stuff like that.
At the end of the day, closer, and these are pretty far off when you're talking about milliseconds. So it's all random. And this was Ted asked, as well, I asked one of the blue that was gone on here. So I looked at that script, broke it down a little bit more.
You see here, it's referenced inside of the scripts, this JavaScript right here. I highligh so you can see it. I open that up. And what we found was their JavaScript was a math is a random number generator. So I was going through and creating all these things and saying is your LCP, this is your time to first buy the FID etc. And it's using funny math to do it.
On top of that, you notice in my original example, there was only 3 metrics that are getting sent to Google. But Google is actually tracking 5, and getting used to send these with a script all into there. So I thought I was kind of shady, if they're going to collect 5 things, and they want you to install this, to give you some more information than they could at least written the code so that you get to see all 5 in your analytics in your log file.
So, I rewrote the script and put it all right there for you. You can drop this on your homepage, and then go to that Google site, it'll tell you how to connect to analytics and how to see this for yourself inside of on your website, or you can just go to digitaleer.com. That's my website and do the log thing. And you can see all that stuff as well. Just make sure again, you have to follow those directions in there so you know how to read the log and get that the fire off.
Here are the test pages I created. I did some different things to them to just kind of mess with those numbers to see if we see some consistency, and or not. And you can see the changes in here. And these are the URLs. I will be publishing this test publicly to kind of promote SIA, as well.
So, these domains, as you are probably familiar, you've seen these domains in other tests. I killed them, all the other tests are all deleted, these pages are set to no index nofollow. And I fully expect to get a manual penalty sooner rather than later when all the narc SEOs go right into Google, like they did to Kyle. So that's why I killed those, I'll delete these in and get some of the test domains. After this is all published. Kill and dropped those domains.
But every one of them, every one of these things that had that script on him, I think Ted removed off of his by now. But every one of them came back with random numbers.
On top of that, when Ted removed that script, that little 5-6 line script, his Core Web Vitals scores went up 30%. So that one little script that they want you to put on your website will actually drop your Core Web Vital scores by 30%. That is just redonkulous, in my opinion. And Ted's site is so sensitive to that, because it's pure HTML.
So it's not a WordPress, and he's kind of CMS, HTML, hard coded that website. And that's why it was significant. So imagine if you put that on a WordPress or something like that, you'd really think with your numbers. And a lot of these tools are running off of the same thing.
So when I say a lot of these tools page, what's it Pingdom GT metrics webpagetest.org. They're all running off the same API. And you can still see the random variants inside of these, the numbers that they're getting.
And what's really really interesting is they're all running on the PageSpeed API. And the PageSpeed API is a average of the last 28 days of Googlebot went to your website. It's not even real time.
You go into these testings tools, you're not even getting the real time data from Core Web Vitals, you're getting the last 28 days, the average of the last 28 days and the metrics. And then there's the Chrome user experience, which has been, which I mentioned, people were saying that user experience is what the real factor is and it's a consolidation. And that creates Core Web Vitals, well, that's updated monthly. And that's what's given to these tools from the PageSpeed API.
So I mean, at this point, we know it's a random number generator, we know the API is based off old data. So how can random numbers be a legitimate ranking experience? And how can random numbers be a legitimate ranking or user experience thing, which is the CrUX data.
Here is a list of all the API data that you get in the API. All of these things, report something in the API. You can pull this out using Screaming Frog, just drop your keyword in there, turn on the PageSpeed indexer API, and it'll go to town and pull all that stuff for you. I did that for all of these keywords.
I downloaded the top 100 URLs that are ranking for it, I used a tool called hexomatic to do that it was stupid easy. And then I put those all in the to the list function in Screaming Frog, which imported the URLs and then got all the API data. And then you have to use the main export button, there's two export buttons, one under the PageSpeed tab, and one under the main one, it's on top of the software. And it'll pull it spit it back out in the order that you put it in, in our case, ranking position 1 to 100.
So, it spit it right back out. And then I use Ted's correlate editing tool to drop everything. I was in spreadsheet hell. It's 100 spreadsheets here, and then another 100 spreadsheet here. And then like 10 spreadsheets to put everything back together.
SoI hope you appreciate the depth of which I had to go through to find out or prove was a factor or not. So in the write up, you're going to see that I broke down these measurements. And then I actually provided a definition, if there was a definition to be had for them. Some of these do not, you just couldn't find them. Some of them were had definitions by Google, others had just kind of educated guesses. And then others, there was nothing.
We kind of just use common sense, rest of them. But as I pulled all this stuff, the tools are all running off of this random math API. And we can see here, performance score, it's not a factor. When you're looking at these charts, right.
The the first image is a specific keyword that I just looked at one keyword, for example, for performance score. And then the second one is where I looked at all 100 or 10 keywords, the top 10 of those to see if there's a ranking.
I could have gotten all 100. But then these charts would be totally unreadable. And it wouldn't change what we're seeing here, because I picked them all randomly, just kind of open the file and boom, got 10. And you see this keyword, top 10, kind of sorta, but the correlations are nowhere clear to be seen significant. We want these bars to turn green.
At the end of the day, we want these lines to go straight. Those two other things happen and we know we have a ranking factor and performance score was not it. So that one's down. First Contentful paint time, FCP, not a factor, not even close, not anywhere in any keyword, it was even close to being a factor.
Speed Index time, how quickly the content is visually displayed during the page load. Seems important. Users want that fast, right? Not a factor. Largers Contentful Paint, measures the time a website takes the show to the user, not a factor. Not even close. Not on any of the keywords. Time to interactive, how long it takes to the page become fully interactive?
Its interactive when its page is shown. There's event handlers and user interactions, like you move your mouse on that page and stuff. Those kinds of fire out those events not a factor. Total Blocking time.
Remember, we remove blocking scripts, eliminate blocking scripts, defer your JavaScript all those recommendations. Yeah, not affect. Cumulative layout shift, not affect. Total requests, reduces the number of HTML requests on your page. You have too many. And this is something that I fixed to this day, right. So to just go to some web pages, and they're trying to load 300 things. I always reduce this. Still not a factor. Total bite size, your entire website, all your stuff.
HTML, CSS JavaScript, not a factor when combined. Weird. Remember this, remember this size, but this one here when you combine it all. Not a factor. Image Size, this is a combination of all your images added up. So if you have 10 images, and they're all 13 gigabytes each, Google did not care, not a factor. Image count, 1 image or 1000 images, doesn't matter.
I mean the high right here, the low is in the teens and the highs are well into the 80s. I think that number is closer to 90, not a factor. CSS Count, count doesn't. But look at what size is doing. Kind of correlating, getting a little closer to a straight line.
We got this weird outlier here in here. But it's kind of correlating. So remember, I told you remember, size, a keyword, JavaScript count? Not so much. Size, a little bit. Not too much, though. Some words, Yes. Some words, No. Font size, they don't care. You load 1 font or 500 fonts, they don't care. Media size, put all the videos, audios all that stuff, Google doesn't care. Other stuff, this is just anything that they didn't factor into the other things right.
So, I don't know, maybe you've got some weird PHP thing going on or something like that, they they lump that into other, doesn't matter.
Third party stuff, this kind of matters a little bit. They're saying don't go crazy loading all those scripts, if you're in love with your chat, and you're embedding 100 maps and doing 50 videos and all that, you might have gone a little nuts, for the top 10.
When you get into the top 100, not so much. And then this is the count, on average of all the keywords. So 3rd party count, not really a factor. I mean, in this case, because we're looking at top 10, maybe you should keep an eye on that. If you're in number 10 and you want to be number 1, look at your 3rd party stuff.
But at the end of the day, if you're competing and you're below 10, then it's one of those things you don't need to worry about. And that's consistent across the board when you're looking at the top 10. So top 10 of 10 keywords you see it's not a factor. 3rd party size, not a factor. Google is not penalizing you for pulling API's.
I used to tell people, hey, you need to reduce that because it's going to hurt you. The correlation say it's actually not. This one is funny, the core web vitals assessment is literally you either pass or you fail. That is the measurement that's returned in the API. Pass Core Web Vitals, or failed Core Web Vitals, I'll tell you, only about a third of the top 100 on all the 76 keywords that I pulled, even have a pass or fail score. 90% of the URLs that are in the top 10 for all 76 of those keywords fail.
Failing the test in the Core Web Vitals assessment is not a ranking factor. I mean, you could fail. I mean, at this point, like Ted suggested, you might just want to fail to get over it. Because you see more sites, you know, they don't pass, here's pass.
Fail is this medium line right here, you see all these sites that are failing. And then these are all the sites that didn't have any test scores. And look, they're failing predominantly. And so that's what I said the Core Web Vitals, you get the humor and stuff if you actually read this yourself. I had the humor so you weren't going to read 43 pages. But enter the content of CrUX, the user experience.
Everyone's pouting off and pouting that saying that user experience is there. Google's measurement of user experiences FCC, FCP, LCP, FID time, TIP, total blocking time the kuna layout shift, we already looked at those numbers, but Google reports them separately out of the cracks report, the CrUX report is updated once a month. And you can see here, first Contentful paint, not a ranking factor.
Give me the layout shifts kind of correlated a little bit. So this one you might want to be careful with. But at the end of the day, when you push all those other ones out, and you look at the different keywords across the board, not a factor. So in this keyword, yeah, you might want to worry about it. But for in general, No, you don't have to worry about Layout shift.
First Contentful type type not even close to being a factor, largest content type not even close. So Chrome user experience is not a factor, user experience. in general is not a factor because these are, this is how they're measuring it. They tell us right there. This is how they're measuring it. Their API is telling us how they're measuring it. And they're not using it. They're not using it as a ranking factor.
But wait, there's more. "Size", remember, I told you to worry about size. Look at this, the HTML size is correlating with higher rankings, in particular in the top 5 of these keywords, right. So if you want to get into the top 5, you need to look at how your website's built, how can you code this, and how big the HTML size is.
When you add up, the slight correlation with JavaScript or and CSS size, we're actually on to something. We're actually on what? Google was using for ranking factor when talking about Core Web Vitals, it was at the Core Web Vitals scores. It was size. We see here HTML size of the ranking position. There's just one outlier right here. It's high but all of the sites in there are high.
So in this case, we can see that size does matter. You want to have a smaller website, what they're trying to do is get you to actually use this as a tiebreaker. So all those people will say, Oh, Core Web Vitals was a tiebreaker or out of something. But it's not Core Web Vitals is tiebreaker. It's your HTML size that's the tiebreaker.
So, if you're in the top 10, and your timesharing is Ted calls it, switching between 2 web, your URL and someone else's your URL, and it keeps going over, go reduce your HTML size a little bit. And you might lock yourself in to that position that your timesharing for.
So why Core Web Vitals? Why Lighthouse, CrUX and all that, at the end of the day, Google wanted you to fix size, your HTML size. And all those eliminate render blocking scripts, you know, defer your loading, using async and or defer and the JavaScript and your CSS, all of that was about reducing your HTML size.
If you reduce your HTML size, Googlebot can crawl your website faster, cost Google less money ergo. Now we have Core Web Vitals and this bill campaign to get you to fix all these things.
When I could have just said it was reduced the HTML size, make your site's cleaner. But most people would never get to the answer of what do I have to clean. So they provided all these tools for us to give us a whole bunch of things to clean in order for us to reduce the HTML size on our websites.
That is the actual ranking factor inside of for PageSpeed optimization. The HTML size of your websites. And that is a month and a half my life that I get back now, or I'm never gonna get to figure that out. But I'm glad I did.
Dori: Wow. I mean, I just incredibly impressed for one thing that you spent a month and a half, and I can empathize with the pain of the spreadsheets and everything. But oh my god, the data that you just found. I mean, this is incredible. Not a factor, not a factor, not a factor. Holy crap. I mean, is this what you're presenting it SEO Rockstars?
Clint: I'm actually I was gonna do schema, but I'm thinking about this one.
Dori: Well, you can do both. I mean, you know, if you do this at SEO Rockstars then let's publish it afterwards. Let's clean up the report and publish it as a white paper, or maybe a shorter article on the site that points to the white paper that people can go read all the data, but oh my god, you know, people are gonna freak out.
Clint: Yeah, for sure. I think this will be a game changer as long as we can get it and get it out there and get in front of people. We do have a year and a half of marketing to kind of undo all these people. The people who are trained to do Core Web Vitals, they all spent a lot of money and now here we are coming along saying hey, you just.
I don't necessarily say they wasted their money. However they were chasing the wrong metric. And I said it in here is everyone that I've talked to you right here, everyone that I talked to you about Core Web Vitals and they say that they got improvements.
When you delve down into the conversation more, what did you do? It always turns out to be, we cleaned up our code, we remove things that we didn't need and our site got faster and our Core Web Vitals went up. So ergo, they reduce the size of their HTML, JavaScript and CSS. And now we just know what Google was after, what they're looking for.
Dori: Wow, it's amazing. And I think it'd be a lot of people are gonna freak out thinking how much money they spent doing some of the stuff and maybe they didn't reduce the size of their HTML, and they still don't really know what's going on why they're not ranking.
Clint: Or those that have good Core Web Vitals scores now, but their sites are still slow.
Dori: Would this pertain to WordPress? When you're talking about HTML, WordPress is extremely heavy in code. So I mean, we've done test, HTML versus WordPress and HTML sites do rank better, In general, but would you advise a client to switch over from WordPress to HTML?
Clint: No, but I would advise them to have their theme reviewed. And if their theme is full of blow, like if, for example, Elementor was really bad at it.
They got better, but they were really bad at putting a lot of code in there and loading, like you had one feature you're using on one page, and that script will load across the entire website. What's a good example, contact. Was that contact form, contact us I think is called or whatever. It's like the one that loads on every page, whether you got a form on there or not.
Gravity Forms does a little bit too. For example, you know, looking at those things and saying what is loading on this page that is not being used, but it's still being loaded. And that's dirty code. And page builders and themes are really bad at that. It's been my experience of WordPress on its own is not necessarily a bad thing. I mean, they're doing some work to reduce the code inside of the WordPress core.
But at the end of the day, I don't think that that is going to have all that much difference for website owners. I think people theme developers are the ones that are really going to need to step up their games. I mean, we all want these fancy shiny things, and fancy shiny things require code, sometimes they require a lot of kid.
We need our devs to not think solely on let's get this built, they need to think more, how do I get this built with the least amount of code possible. And that's where the game is gonna win.
Dori: Do you know of a place where people can test themes for? This kind of thing?
Clint: Well, because we know that its a random number generator now. I mean, you can kind of compare a little bit using the testing tools. But at the end of the day, it's really you're just going to have to actually look at, I don't know, maybe even putting a bunch of themes up on some domains and then running a tool like Cora or something like that. That's actually going to count the size of your HTML files for you.
Dori: Yeah, that would be interesting.
Clint: And what are you really looking for at the end of the day is something like this SEO, something that's going to tell you this, the size of this slide. This is all the stuff that's in the DOM. This is all the HTML file is loading for this particular website.
You can see even on SEOIntel, the theme is got a lot crap, right? So all this CSS and all this stuff, it's all populating and it's all creating stuff. So what is the actual bite size of that? That's being out so you need tools in the market, they'll give us that and that'll probably help us out significantly.
If you want to chase the HTML size rabbit. I wouldn't get into that, unless of course you're stuck in the top 5 position you want to rank out, you've done everything and you're looking for something to kind of break the tie, HTML size might be the way to go now.
Dori: Yeah, well there's a lot of people out there in that position, I think.
Clint: Yeah, for sure.
Dori: Or looking to get into the top 3 instead of 5 and below, above the fold. And this would be a definitely an above the fold type of move.
Clint: Especially if you've done everything else. I mean, there's always something else you can do. But if you've done everything else, and you got a dev who's got some time in their hands, point him in this direction, tell him to clean up the code.
Dori: Wow Clint, that was awesome!
Clint: I did my little clappy function, but I don't think it works in here.
Dori: I don't think it works in GoToWebinar unfortunately, I think it's just a YouTube thing.
Dori: Well there's my claps.
Clint: Cool
Dori: Okay, we're over the hour. I don't think we have any questions on that. So you guys were the first to find out. The first to know.
Clint: Well, I gave this to IMG as well, because, you know, Ted and Lee inspired this, but they got this today, too. So all you guys know, today, I think there's just about as many people in there so now there's 10 SEOs that know the truth.
Dori: Are they gonna publish it as well?
Clint: They have a PDF. They have this branded SIA PDF but they don't actually have the like, it's not posted. You know what I mean?
Dori: Okay.
Clint: Like I said I wanted to do this publicly, push this out and kind of drive, uses data to promote Digitaleer, SIA and markets with this stuff.
Dori: I'll publish it to the list and say, Whoa, we just broke, you know, huge myths on this stuff and got down to the core of the Core Web Vitals.
Clint: Yeah, for sure. Hopefully, I didn't oversell it. I know, I was talking to a lot of smack in the Skype group and stuff about, hey come up with some new data and stuff. I mean, obviously, everyone knew my position that Core Web Vitals, I didn't think they should be optimizing for Core Web Vitals. So it wasn't an entire secret was coming.
But I think, using all this data that kind of definitively put this to bed. And it did allow me to pinpoint the actual ranking factor inside of here that Google.
Dori: Okay, awesome. I really enjoyed that, actually. So any questions on that? None is coming in.
Clint: Yeah, well, it is a lot. So I'll say go back, read the PDF. I did throw a lot of humour in there. So that it makes it more easier to digest. Again, some of the descriptions and stuff are in there to kind of help you filter that stuff out and understand exactly what they're talking about. But the too long didn't read version is reduced HTML size.
Dori: And the work that you put into that is just overwhelming. Wow, I mean, it's incredible.
Clint: Yeah, it was quite the rabbit hole. I'm not gonna lie.
Dori: I hope you had a little bit of fun.
Clint: Yeah.